Research on Master-Slave Game Optimization Scheduling Strategy for Virtual Power Plants Incorporating Photovoltaic and Thermal Storage Electric Heating Clusters

In order to maximise the potential flexibility of winter electric heating loads, this study proposes a master-slave game-theoretic optimisation scheduling strategy for virtual power plants incorporating a photovoltaic and thermal storage electric heating cluster. This strategy is formulated as a leader-follower game with dynamic clustering. First, a dynamic clustering approach based on the K-means algorithm is employed to analyse the load characteristics of differentiated users, taking into account dynamic factors such as the charging/discharging power of thermal storage devices and variations in indoor temperature. Next, a leader-follower game optimisation model is established in which the VPP acts as the leader coordinating the photovoltaic-integrated thermal storage heating cluster. Given the differing economic objectives of the VPP and its users, the upper-level model aims to maximise the VPP’s operational revenue, while the lower-level model seeks to minimise users’ heating costs. Finally, Karush–Kuhn–Tucker (KKT) conditions are applied to transform the nonlinear, two-level model into a single-level, mixed-integer linear programming (MILP) problem. The results demonstrate that the proposed strategy can effectively meet users’ heating demands while ensuring the economic benefits for both parties, thereby achieving a mutually beneficial outcome for the VPP and electric heating users.